David J. Finkel

Department of Anthropology, Adelphi University, Garden City, New York 11530, U.S.A.

Received 19 July 1976 and accepted 24 October 1977

Spatial and Temporal Dimensions of Middle Eastern Skeletal Populations

Osteometric variables of 48 local populations from Southern Europe and the Middle East, ranging in time from 3100 B.C. to 200 A.D., are analyzed to determine their spatial and temporal relationships. Probit analysis is carried out on each population to insure the norm- ality of each sample. The Student-Newman-Keuls test is used to determine significant differences between population means for populations which vary in time but not in space. The few significant differences which exist can be explained by either drift, selection or migration. The population means of all variables are plotted against time to determine whether significant correlations exist. Only two of 16 variables show a significant correlation with time. A model ofthls region as a single interbreeding group of populations is rejected.

F-ratios and Bartlett's test are used to test models in which at least partially isolated subgroups are assumed. A model based on modern political boundaries is rejected because of non-significant F-ratios and heterogeneous variance. However, a model based on spatial isolation and ancient cultural boundaries shows significant F-ratlos for all variables except one; also, Bartlett's test confirms the homogeneity of variance for all variables. In the evolution of these populations during the time period under consideration, it is concluded that evolution proceeds slowly; small changes per generation of phenotypic character- istics are noted. Except in a few cases, migration is believed to have little effect on the population evolution of this area. Drift also is believed to have little effect, so that the primary factor affecting these populations is natural selection.

1. I n t r o d u c t i o n

I f it may be assumed that h u m a n evolution has been an on-going process since the end of the last glaciation, the invest igat ion of recent skeletal populat ions ma y reveal some of the phases of evolution and differentiat ion as modern h u m a n populat ions adap ted to post-glacial env i ronmenta l conditions.

Biological relationships between populat ions are of interest as possible indicators of in te rbreeding or the degree of isolation between populat ions. The presence or absence of significant in te rpopula t ion var ia t ion may also be of use in establishing eco-biological and, by inference, cul tural boundar ies over a larger geographical area t han is normal ly impl ied by the term Mendelian population. For example, Birdsell (1950) proposes the use of isophenes to discover the boundar ies between Mende l i an populat ions and to measure biological distance as an indica t ion of popula t ion divergence. However, because of gaps in the fossil record, both spatial ly and temporally, it would seem difficult to establish the presence of isophenes or clines for a var iable a l though observed var ia t ion be tween populat ions may suggest the boundar ies of Me nde l i a n populations.

Wr igh t (1931) has proposed a model of evolut ion which appears to be most appl icable to man , i.e. small, par t ia l ly isolated subpopulat ions. I t is impor t an t in the analysis of the fossil populat ions of a geographical region to de termine if Wright ' s (1931) model is appropr ia te or if a large panmic t ic popula t ion model or a small, completely isolated popula t ion model is more successful.

Journal of Human Evolution (1978) 7, 217-229 0047-2484[78/0301-0217502.00/0 �9 1978 Academic Press Inc. (London) Limited

218 D.J. FINKEL

In order to determine this, it is necessary to estimate both the amount of change taking place in Mendelian populations over time and the extent of interbreeding between syn- chronic populations. However, the very nature of skeletal populations requires certain restrictions, as follows.

(1) Partially isolated groups with partially restricted gene flow within the group is assumed, i.e. non-random mating behavior.

(2) Although mutation and recombination are sources of variation, they will be assumed to be constants in the Mendelian populations of this study. Since the time period under consideration here does not exceed 200 generations, it does not appear likely that a non-lethal mutation or recombination would significantly affect the Mendel- ian populations of this study. Practically, the identification of a lethal allele in a skeletal population would appear to be difficult, at best.

Significant factors affecting the Mendelian or local populations of this study include selection, migration and drift. The relationship between migration and drift is of import- ance since Wright (1931) has stated that even gene flow of considerable magnitude may be insufficient to prevent the effects of genetic drift. In addition, differentiation between local populations may be adaptive (due to selection pressure) or non-adaptive (due to drift). Therefore, interpopulation variation will be assumed to be the result of the interaction of selection, drift and migration.

Because of the large amount of skeletal material available and the numerous public- ations containing biological data on skeletal populations, the Middle East, Greece and Italy were selected as the area for study. A total of 48 local populations contain sufficient skeletal material for analysis. The time range of these local populations extends from 3100 B.C. to 200 A.D. Table 1 lists the local populations, the sample size and the approx- imate temporal and spatial boundary of each local population. It may be noted that there are 32 archaeological sites, seven of which yield local populations from different time levels. References are to the dating of skeletal populations; whenever possible, the original material was measured by the author. Where a local population has been dated approximately, i.e. a range exists in the dating of the level, the midpoint of the range is used in correlating the local population with absolute time.

Table 1 Local p o p u l a t i o n s in t i m e and space

Site (N) Dating Source

Anatolia Alacahuyuk ( 1 3 ) 2500-2300 B.C. Kansu (1937) Allsharhuyuk a (18) c. 2300 B . C . Krogman (1933, 1937)

b (16) 1000-800 B.C. c (18) 500-300 B.C.

Kultepe (15) c. 1900 B . C . Senyurek (1952)

Cyprus Lapethos a (19) 2500 B.C. Buxton (1920)

b ( 5 0 ) 1800-1600 B.C. c ( 1 3 ) 1200-1000 B.C. d (16) 300-100 B.C.

Sotira (62) 3200-3000 B.C. Angel (1953)

DIMENSIONS OF M. E. SKELETAL POPULATIONS 219

Table I (conr

Egypt Abydos (48 c. 2900 B.C. Sekkara (44) c. 2900 B.C. Sedment (71) c. 2100 B.C. Thebes (55) c. 2000 B.C.

Gree~ and Cre~ Athens a (8) c. 2600 B.C.

b (31) c. 2400 B.C. c (26) 1400-1200 B.C. d (52) 1100-1000 B.C. e (33) 900-700 B.C. f (15) 500-300 B.C. g (50) 300-100 B.C.

Olynthus (16) c. 400 B.C. Zakro (22) c. 1500 B.C.

Morant (1925) Bat rawi&Morant (1947) Woo (1930) Batrawi&Morant(1947)

Angel (1944, 1945, 1946)

Angel (1942) Dawkins (1900-01)

Iran Shah Tepe (12) c. 1800 B.C. Furst (1939) Tepe Hissar (336) c. 2000 B.C. Krogman (1940) Tureng Tepe (18) c. 2000 B.C. Deshayes (1963, 1973)

Iraq Kisha (47) c. 2500 B.C. Buxton & Rice (1931)

b (60) c. 500 B.C. Nippur (32) 700-500 B.C. Swindler (19565 Nuzi (46) c. 200 A.D. Ehrich (1939)

/sTeel Gezer (88) c. 1450 B.C. Lachish a (79) 2400-2200 B.C.

b (863) c. 700 B.C. Megiddo a (18) c. 2900 B.C.

b (14) 1600-1400 B.C.

Finkel (1974) Risdon (1939)

Hrdlicka (1940)

/ta~ Etruria (28) 700-500 B.C. Sergi (1915) Marsiliana (8) c. 700 B.C. Ciprlani (1927) Rome (70) c. 100 A.D. Brothwell (pets. comm.)

Syria Chatal Huyuk (18) 1400-1200 B.C. Krogman (1949) Palmyra (30) 100-300 A.D. Brothwell (pers. comm.) Tell-al-Judaidah (24) 1400-1200 B.C. Krogman (1949)

Coastal Turkey Bodrum (14) 200 A.D. Tunakan (1964) Ephesus (74) 300-100 B.C. Schumacher (1926) Karatash (142) c. 2400 B.C. Angel (1966) Sardis (86) c. I00 A.D. Bostanci (1963, 1967,

1969) Troy a (26) c. 2600 B.C. Angel (1951)

b (25) c. 1500 B.C. c (28) e. 500 B.C.

A n o t h e r r eason for t h e se lec t ion o f the M i d d l e Eas t as t he a r e a for s t udy is t he l a rge

a m o u n t o f his tor ical , l inguis t ic a n d a r c h a e o l o g i c a l i nves t i ga t i on ca r r i ed o u t in this r eg ion .

Ex i s t i ng cu l tu ra l b o u n d a r i e s p r o v i d e a c o n v e n i e n t hypothes i s for tes t ing the b o u n d a r i e s

o f g e n e f low a n d the a m o u n t o f c h a n g e w h i c h has o c c u r r e d in a l l och ron ic , s y m p a t r i c

220 D.J. FINKEL

populations. Intrapopulat ion variation and interpopulation variation will be used to infer the combined or individual effects of migration, drift and selection on the local populations of the Middle East and Southern Europe.

2. M a t e r i a l s a n d M e t h o d s

In the description of human populations of the Middle East, nearly all the biological data collected consist of osteometric data, especially of the cranium and mandible. In order to obtain sufficient statistical samples, these are selected as the material for study. Further, since the greatest proportion of individuals available are males (in fact, some populations consist only of males), males only are used.

Although precise genotype-phenotype relationships are difficult, if not impossible, to confirm for quantitative variables generally used in the study of skeletal populations, certain of these variables may allow inferences to be made concerning the statistical differentiation of populations and may provide a standard by which the evolution of a single local population may be measured.

Osteometric variables of the cranium and mandible were selected to reflect the dimen- sions and structure of the cranial vault, the base and height of the cranium, the face and the mandible. The 12 variables of the cranium and four of the mandible are presented in Table 2.

Table 2 O s t e o m e t r i c v a r i a b l e s o f the c r a n i u m and m a n d i b l e

Measurement (Martin's number) Variable number

Maximum length (I) 1 Maximum width (8) 2 Basion-prosthion (40) 3 Biaurieular breadth (11) 4 Basion-begma (17) 5 Basion~nasion (5) 6 Nasion-prosthion (48) 7 Bizygomatie diameter (45) 8 Nasal height (55) 9 Orbital breadth (51) 10 Orbital height (52) 11 Interorbital breadth (50) 12 Bigonal diameter (66) 13 Condylo-symphyseal length (68) 14 Symphyseal height (69) 15 Ascending ramus breadth (71) 16

Each local population is first tested for homogeneity. I t is possible that a skeletal population may actually be composed of two or more groups because of disruptive selection, because interbreeding has been restricted by cultural barriers, or because groups from different areas have only recently merged. Homogenei ty is established through the use of probit analysis. When the observations for one variable of the local population are plotted against previously assigned probits, a linear regression line indicates a normal distribution (Finney, 1971). I t is suggested that a population which is non-normal is likewise heterogeneous in the sense that it is not in equilibrium because

D I M E N S I O N S O F M. E . S K E L E T A L P O P U L A T I O N S 221

of migration, selection, inbreeding, or because the obtained sample does not accurately represent the original population. Probit analysis of the 48 local populations of this study (not presented because of space limitations) indicates no significant departure from normality for any local population.

The use of multivariate analysis in the study of population dynamics, especially to determine genetic distance between populations is widely known (cf. Barnard, 1935, Crichton, 1966; Giles & Elliott, 1963; Howells, 1966; Talbot & Mulhall, 1962). How- ever, nearly all existing multivariate studies have been carried out on one or, at most, several populations where a complete series of measurements is available. Multivariate analysis of a population in which many of the individual measurements are missing is very difficult. The use of many populations with widely disparate sample sizes raises other problems, for example, the weighting of populations. Also, Knussman (1968) found that the Penrose distance technique did not correlate with the Mahalanobis 132 in multivariate analysis of the same populations when sample sizes were small. Finally, the investigation of the dynamics of many local populations may be successfully carried out using alternative techniques of univariate analysis.

In order to establish spatial boundaries for the 48 local populations, the extent of differences which exist between local populations must be determined. Obviously, the largest possible group is one which contains all 48 populations forming a large inter- connecting gene pool. The smallest spatial and temporal grouping consists of each local population as a separate entity, perhaps partially or even completely genetically isolated from the other populations.

An acceptable basis for the grouping of local populations is one in which the variation within a subgroup is statistically significantly smaller than the variation which exists between the subgroups. In order for the model to be statistically significant rather than fortuitously so, only one variable of the 16 variables is used to divide the populations into subgroups. Then, the remaining 15 variables are examined to determine whether they are significant as well. This eliminates the danger of a spuriously significant model. Generally, the statistic used to determine the significance of intrapopulation variance and population variance is the F-ratio (Sokal & Rohlf, 1969) ; the 0.05 level is used to deter- mine significance.

To confirm further the model, Bartlett's Test of Homogeneity of Variance is carried out on the subgroups for each variable. This test determines whether significant difference in the variances of the different subgroups exists, or whether the variances of each sub- group are approximately equal. Heterogeneous variance might lead to a coincidentally significant model so that any significant heterogeneity (at the 0-05 level of significance) would tend to disqualify the variable from consideration. Several heterogeneous variables might cast doubt on the validity of the model itself.

Additionally, the population data present the opportunity to eliminate the spatial variable and to examine population evolution taking place in a single location. Two or more local populations may be derived from a single archaeological site. The determin- ation of significant difference between such populations is accomplished through the use of the Student-Newman-Keuls Test. This test is similar to the Student t-Test, but the advantage of this test is that it may be carried out simultaneously on all the local popul- ations found at the site despite the fact that sample sizes may be unequal (Sokal & Rohlf, 1969). The rate of change per generation may also be calculated for each variable that shows significance, assuming a 20-year generation span.

222 D.J. FINKEL

I n o r d e r to cons ider t h e e v o l u t i o n o f al l t h e loca l popu l a t i ons , t he m e a n s o f p o p u l a t i o n

va r i ab le s m a y be p l o t t e d aga ins t t i m e w h e r e the t i m e r a n g e o f the p o p u l a t i o n s o f this

s t u d y ex tends f r o m 3100 B.C. to 200 A . D . T h i s m a y d e m o n s t r a t e the p r e sence o r absence

o f i nc rea s ing or d e c r e a s i n g t r ends in t h e v a r i a b l e o v e r t ime , us ing the P e a r s o n ' s r cor re l -

a t i on coeff ic ient (s igni f icance is a t t he 0.05 level) .

3. D i s c u s s i o n a n d C o n c l u s i o n s

T a b l e 3 lists the loca l p o p u l a t i o n s w h e r e t h e spa t ia l f ac to r has b e e n e l i m i n a t e d . Also

d e s c r i b e d in the t ab le is the n u m b e r o f t i m e levels p re sen t a t e a c h a r c h a e o l o g i c a l si te as

we l l as t he a m o u n t o f t i m e b e t w e e n consecu t ive levels. T h i s last s ta t is t ic is c o n v e r t e d in to

g e n e r a t i o n s w h e r e a t i m e span o f 20 years p e r g e n e r a t i o n is a s sumed .

T a b l e 3 Local p o p u l a t i o n s f r o m archaeo log ica l s i t e s

Site Level Time difference Number of generations (year)

Alisharhuyuk a a - -b = 1400 70 b b - -c = 500 25 c a - - c = 1900 95

Athens a a - -b = 200 I0 b b - -c = 1100 55 c c - -d = 250 12-5 d d - -e = 250 12.5 e e - - f = 400 20 f f - -g = 200 10 g a - -g = 2400 120

Kish a a - -b = 2000 100 b

Lachish

Lapethos

a a - - b = 1600 80 b

a a - -b = 800 40 b b - -c = 600 30 c c - -d = 900 45 d a - -d = 2300 115

Megiddo a a - -b = 1500 75 b

Troy a a - -b = 1100 55 b b - -c = 1000 50 c a - -c = 2100 105

T h e local p o p u l a t i o n s f r o m a n a r c h a e o l o g i c a l site a r e t h e n tes ted for s ta t i s t ica l signifi-

c a n c e o f d i f ferences b e t w e e n means , u t i l i z ing the S t u d e n t - N e w m a n - K e u l s test. T h e

resul ts o f these tests a r e p r e s e n t e d in T a b l e 4(a) .

T h e r e a r e four sites ( i nc lud ing M e g i d d o w h e r e o n l y one v a r i a b l e cou ld be tested)

w h e r e abso lu te ly no s ta t is t ical s ign i f icance exists b e t w e e n p o p u l a t i o n m e a n s for a n y o f t he

Table 4

DIMENSIONS OF M. w. SKELETAL POPULATIONS 223

(a) Significance of differences between local populations

Number of Number of Number of Site levels variables tested comparisons

Number-percentage of significant comparisons at 5% level of

significance

Alisharhuyuk 3 12 36 0 Athens 7 15 315 4-1'27 % Kish 2 12 12 0 Lachish 2 12 12 3-25"00% Lapethos 4 12 72 2-2.78% Megiddo 2 1 1 0 Troy 3 13 39 0

(b) Variables and populations showing significant difference

Difference in population Number of Change per Site Variable means (mm) generations generation (ram)

Athens b- -e Orbital breadth 3"29 80 0.041 Athens b-- e Ascending ramus 4.70 80 0.059

breadth Athens b-- g Orbital breadth 3.15 110 0"029 Athens d-- -- g Nasion prosthion 4"38 42"5 0" 103

Lachish a - -b Maximum length 3.07 80 0.038 Lachish a-- b Biauricular 4-56 80 0"057

breadth Laehish a-- b Orbital breadth 2"17 80 0"027

Lapethos b - -d Basion nasion Lapethos b-- d Nasion prosthion

6.50 75 0.087 4.32 75 0.058

var iables tested. I t m a y be inferred, therefore, tha t selection, migra t ion a n d drif t have not p l a y e d a signif icant role in the evolut ion of these popula t ions in the t ime span in- volved, and the differences wh ich exist be tween the local popula t ions m a y be asc r ibed to s ampl ing error a n d na tu r a l selection.

T h e significant differences discovered in the popula t ions of Athens, Lape thos a n d Lach i sh are broken down in T a b l e 4 to i l lust ra te the var iables . O f the tests for s ignif icance in the popula t ions f rom Athens only 1.27 % in fact showed signif icant differences. Th is suggests tha t the site was s table (in terms of genet ic equi l ib r ium) dur ing the pe r iod u n d e r considera t ion. T h e largest ra te of change per genera t ion is found be tween Athens popula t ions d a n d g. I t is l ikely tha t the change reflects the increas ing me t ropo l i t a n na tu r e of the a rea as well as a c c u m u l a t e d dr if t a n d selection.

T h e Lape thos popula t ions also show an insignif icant percen tage of s ignif icant differ- ences be tween local popula t ions (2 .78%). I t m a y also be no ted tha t no s ignif icant differences exist be tween t ime per iods b a n d c or be tween c and d ; yet, there a re differ-

224 D . j . FINKEL

ences be tween popula t ions b and d. These differences are most l ikely the resul t of accumu- l a t e d selection a n d drift .

T h e Lachish popula t ions show a signif icant pe rcen tage of s ignif icant differences (25"00~) . Cons ider ing the a m o u n t of known a rchaeo log ica l and his tor ical mig ra t ion t ha t has taken p lace in this a r ea (Kenyon , 1960), the presence of s ignif icant differences be tween the popula t ions because of m i g r a t i o n is not surprising.

Table 5 Signif icance of o s t eometr i e var iab le s

Variable m r 0.05 0-01

Maximum length --7-026 • 10 -5 --0.017 0.273 0.354 Maximum width 2"523 X 10 -s 0.008 0"273 0.354 Basion prosthion 2"729 • 10 -4 0.097 0.355 0"456 Biauricular breadth 1-253 • 10 -4 0.314 0.444 0.561 Basion bregma 6'212 x 10 -s 0.019 0"304 0"393 Basion nasion 4.602 X 10 -4 0.174 0.304 0.393 Nasion prosthion --4.993 • 10 -4 --0.157 0.304 0'393 Bizygomatic diameter 1-166 X 10 -4 0.384 0.325 0"418 Nasal height -- 1"222 x 10 -5 --0"006 0.304 0.393 Orbital breadth 3"672 X 10 -4 0.233 0"304 0'393 Orbital height 1"521 x 10 -~ 0"136 0.304 0.393 Interorbital breadth --2.466 x 10 -4 --0'174 0.349 0.449 Bigonial diameter 4"263 x 10 -4 0.116 0.423 0"537 Condylo-symphysed length 1.194 x 10 -4 0"306 0.433 0.549 Symphysed height --8"295 X I0 -4 0.456 0"404 0.515 Ascending ramus breadth --1.171 x 10 .5 --0.081 0.413 0"526

T h e m a g n i t u d e of al l popu la t ion means m a y be p lo t t ed agains t t ime in o rde r to infer the presence or absence of t rends in the region as a whole. Graphs , not p resen ted here, suggest the absence of a s ignif icant p a t t e r n across t ime for each var iab le . T a b l e 5 presents the slope (m) a n d corre la t ion coefficient (r) for each var iab le cor re la ted wi th t ime. O n l y two var iables , b i zygomat i c d i a m e t e r and symphysea l height , show signif icant r a t the 0.05 level of significance. Bizygomat ic d i ame te r shows a t endency to increase over t ime whi le symphyseal he ight shows a t endency to decrease over t ime. However , a t empora l model , based on one var iab le , symphysea l height , is co r robora t ed by only one o f the r e m a i n i n g 15 var iables . W h i l e it is possible tha t selection is opera t ing on jus t these componen t s of the skull, the t empora l mode l must be r ega rded as unsuccessful. I t wou ld seem tha t the M i d d l e Eas t du r ing the t ime pe r iod unde r cons idera t ion canno t be r e g a r d e d as a single in t e rb reed ing popu la t ion bu t may , perhaps , be d iv ided into groups of local popula t ions in which each g roup behaves as an in t e rb reed ing unit . I t is assumed tha t the absence o f s tat is t ical ly signif icant differences between the local popula t ions compos ing a g roup suggests tha t these popula t ions , t hough sepa ra t ed in space a n d / o r t ime, m a y be considered to be pa r t o f an in t e rb reed ing group. I f the popula t ions are sepa ra t ed in t ime, i t is impl ied tha t they form an in t e rb reed ing cont inuum.

Large ly to demons t r a t e the ab i l i ty o f the me thodo logy to re ject a fortuitous model , the first mode l is based p r i m a r i l y on m o d e r n pol i t ical boundar ies . T h e d i s t r ibu t ion of popula t ions for M o d e l I is presented in T a b l e 6. T h e results o f F - r a t i o tests a n d Bar t le t t ' s Tes t for homogene i ty of va r i ance for each of the 16 var iables is presented in T a b l e 7 .

T a b l e 6

DIMENSIONS OF M. E. SKELETAL POPULATIONS

C o m p o s i t i o n o f M o d e l I

Group 1-- Turkey Group 5--Italy Alacahuyuk Etruria Alisharhuyuk--3 levels Marsiliana Bodrum Rome Chatal Huyuk Ephesus Group 6--Nile Valley Karatash Abydos Kultepe Sakkara Sardis Sedment Tell-al-Judaidah Thebes Troy--3 levels

Group 7--Iraq (Tigris-Euphrates Group 2--Cyprus Valley) Lapethos--4 levels Kish---2 levels Sotira Nippur

Nuzi

Group 3--Israel Group 8--North Iran Gezer Shah Tepe Lachish--2 levels Tepe Hissar Megiddo--2 levels Tureng Tepe Palmyra

Group 4--Greece Athens--7 levels Olynthus Zakro

225

T a b l e 7 R e s u l t s o f s t a t i s t i c a l t e s t s o n M o d e l I

Bartlett's Variable F-ratio p test p

1 1.36 p =0"25 20"04 0.005 <p <0"01 2 3"22 0"01 <p <0"05 12"60 0"05 <p <0"1 3 1.27 0"25 <p <0"50 11.I0 0,1 <p <0"5 4 2.75 0.05 <p 0-1 10"17 0"1 <p <0-5 5 3.14 0'01 <p <0"025 9"22 0-1 <p <0,5 6 1.30 0.25 <p <0.5 10-44 0.1 <p <0-5 7 6.42 0.001 <p <0'005 13"83 0.05 <p <0.I 8 1.61 0.1 <p <0"25 23.47 p <0.005 9 3.65 0"005 <p <0"01 9-78 0"1 <p <0.5

10 3.35 0.005 <p <0.01 5.26 0.5 <p <0.9 11 2.93 0"01 <p <0"025 9.79 0.1 <p <0.05 12 2-41 0'05 <p <0-I 6-84 0.1 <p <0.5 13 2-33 0.1 <p <0"25 1.87 0"5 <p <0"9 14 5.00 0.01 <p <0"025 5"30 0'I <p <0"5 15 2.32 0.25 <p <0.5 1.62 0.5 <p <0.9 16 2.31 0.I <p <0.25 3"65 0.1 <p <0.5

T h e results show tha t two of the variables tested exhibi t significantly heterogeneous

var iance , and only seven of the variables exhibi t significant F-ratios at the 0.05 level of

significance. As expected, this mode l must be rejected.

I n the construction of Mode l I and the ca lcula t ion of F-ratlos, it was observed tha t

several of the eight groups appea red to exhibi t significant F-ratios while others did not.

226 D.J. FINKEL

I t was also observed that several of the groups, based on modern political boundaries, coincide with ancient political boundaries as well. A second model was constructed based on the following principles.

(1) Populations closest together in space are most likely to be interbreeding units (Birdsell, 1950).

(2) Ancient cultural boundaries, confirmed through historic, linguistic and archaeo- logical sources, are likely to be interbreeding boundaries as well.

(3) Local populations of this study are more likely to vary in space than in time.

In order to prevent the construction of a coincidentally successful model, a single variable was selected in the establishment of the model. The remaining 15 variables were tested to determine whether they satisfied the model as well. The variable selected was max imum length since it was present in greatest frequency in all of the populations.

A consideration of spatial clustering led to a break up of the Turkish group of Model I into two groups: one group consisting of lowland coastal or near coastal sites, the other group composed of Anatolian plateau or mountainous sites. The two different environ- ments of the Turkish sites as well as geographical distance are used to determine the composition of groups. The third group was composed of two populations, one in Southern Cyprus and the other directly across the Mediterranean in Syria. Similarity of group means for Max imum Length as well as archaeological similarities led to this grouping (MacEwan, 1937). The populations from Palestine, Greece, Italy, the Nile Valley, the Tigris-Euphrates Valley and the Iranian plateau were grouped as in Model I. Table 8 presents the composition of groups for Model I I , and Table 9 presents the F-ratio and Bartlett's test for all variables for Model II .

Table 8 C o m p o s i t i o n of Model II

Group 1--Coastal Turkey and Group 5--Greece North Cyprus Athens--7 levels Bodrum Olynthus Ephesus Zakro Karatash Lapethos--4 levels Group 6--Italy Sardis Etruria Troy--3 levels Marsiliana

Rome Group 2--Anatolia and Taurus Mountains Group 7--Nile Valley Alacahuyuk Abydos Alisharhuyuk--3 levels Sakkara Chatal Huyuk Sedment Kultepe Thebes

Group 3 Group 8--Iraq (Tigris-Euphrates Sotira Valley Tell-al-Judaidah Kish--2 levels

Nippur Group 4--Israel Nuzl Gezer Lachish--2 levels Group 9--North Iran Megiddo--2 levels Shah Tepe Palmyra Tepe Hissar

Tureng Tepe

DIMI~NSIONS OF M. lg. SKELETAL POPULATIONS 227

T a b l e 9 R e s u l t s o f s t a t i s t i c a l t e s t s o n M o d e l II

Bartlett's Variable F-ratio p test p

1 2.85 0.01 <p <0.025 6-12 0-5 <p <0"9 2 8.23 p <0'001 13"10 0.05 <p <0-1 3 4-89 0.001 <p <0.005 12,02 0.05 <p <0-1 4 5-34 0"005 <p <0-01 1-41 0"9 < p <0.975 5 2.48 0'025 <p <0.05 9.73 0.I <p <0-5 6 4-05 0"001 <p <0"005 8-66 0.1 < p <0-5 7 6.84 p <0-00t 11-84 0'1 <p <0-5 8 3.36 0"01 <p <0"025 8"37 0"1 <p <0.5 9 3.21 0-005 <p <0.01 9.45 0"1 <p <0"5

10 3.37 0"005 <p <0"01 5-53 0.5 <p <0.9 I1 1-96 0.05 <p <0.1 11-01 0"1 <p <0.5 12 2.46 0'025 <p <0"05 11.40 0,05 <p <0-1 13 4"62 0"025 <p <0-05 0.34 0"9 <p <0.975 14 17.62 p <0"001 0,41 0"9 <p <0-975 15 3.66 0.025 <p <0"05 4.93 0.1 <p <0.5 16 3.76 0.025 <p <0.05 7.42 0.1 <p <0-5

This is a far more satisfactory model since all variables but one show significant F- ratios. The one exception, orbital weight, would be significant at the 0-05 level with an F-ratio of 2-18, so this variable barely fails to show significance as well (1-96). All variables show homogeneous variance. Since the model was formulated from a consider- ation of only one variable, and it is observed that the remaining variables satisfy the conditions of the model as well, Model II may be accepted as a valid combination of local populations into interbreeding groups. Each group may include populations separated in time.

Beyond the boundaries of the group, the phenotypes of the local populations are statistically significantly different from those within the group. Obviously, population divergence has taken place between groups; each group may, in a sense, be regarded as a local population where selective breeding has not caused divergence within the group.

The role of migration, drift and selection in the population evolution of this region may be summarized. Migration appears not to have been sufficient to prevent divergence among the sub-regions of the Middle East. This possibility was predicted theoretically by Wright (1931). I f genetic drift were a major factor in the evolution of these popula- tions, major fluctuations would be expected between local populations. This would be confirmed by comparison of local populations within the same site. As described above (Tables 3 and 4), these fluctuations are rare. The random nature of the osteometric variables graphed against time (Table 5) suggests that drift may have been a minor factor in population evolution, but it appears doubtful that it was a major one.

Selection appears to be the major factor in population evolution. When calculable, the rate of evolution per generation does not indicate rapid evolution. In addition, the normal distribution of local populations confirms that genetic equilibrium was rarely disturbed. Disturbances generally result from migration rather than selection; actually, the rate of evolution due to selection would be expected to be fairly slow since the environ- ment changes very little in the region as a whole in the time period under consideration.

228 D . j . FINKEL

T h e feasibi l i ty o f c o m b i n i n g loca l p o p u l a t i o n s to f o r m i n t e r b r e e d i n g uni ts w h e n the

p o p u l a t i o n s c o m p o s i n g the uni ts a re s e p a r a t e d by long t i m e in t e rva l s f u r t h e r suggests

t h a t e v o l u t i o n is p r o c e e d i n g a t a s low ra te .

R e f e r e n c e s

Angel, J. L. (1942). Classical Olynthians. In (D. M. Robinson) Neerolynthia, Excavations at Olynthus II, 211-240.

Angel, J. L. (1944). A racial analysis of the ancient Greeks : an essay on the use of morphological types. American Journal of Physical Anthropology 2, 329-376.

Angel, J. L. (1945). Skeletal material from Attica. Hesperia 14, 279-363. Angel, J. L. (1946). Social biology of Greek culture growth. American Anthropologist 48, 493-533. Angel, J. L. (1951). Troy, the Human Remains. Supplemental Monograph Number 1. Princeton: Princeton

University Press. Angel, J, L. (1953). Neolithic crania from Sotira. I n (P. Dikaios) Excavations at Sotira. Museum Mono-

graphs. Philadelphia: University of Pennsylvani a Press. Angel, J . L. (1966). Human remains at Karatash. In (M.J . Mellink) Excavations at Karatash-Semayuk

in Lyeia, 1965. American Journal of Archaeology 72, 260-263. Barnard, M. (1935). The secular variations of skull characters in four series of Egyptian skulls. Annals of

Eugenics 6, 352-371. Batrawi, A. & Morant, G. M. (1947). A study of First Dynasty Egyptian skulls from Sakkara and of an

Eleventh Dynasty series from Thebes. Biometrika 34, 18-27. Birdsell, J. R. (1950). Some implications in the geneic concept of race in terms of spatial analysis. Cold

Spring Harbor Symposia on Quantitative Biology 15, 259-314. Bostanci, E. (1963). An examination of some human skeletal remains from the Sardis excavations.

Antropoloji 1, 17-36. Bostanci, E. (1967). Morphological and biometrical examination of some skulls from the Sardis excav-

ations. BeUeten, Turk Tarih Kurumu 31, 1-48. Bostanci, E. (1969). Sardis kazarilinda eikan kafatashlarin incelenmesi ve eski Anadolu halklari ile olan

munasebetleri. Ankara Universitesi Dil ve Tarih-Cografya Fakultesi 185, 1-98. Buxton, L. H. D. (1920). The anthropology of Cyprus. Journal of the Royal Anthropological Institute 50,

183-235. Buxton, L. H. D. and Rice, D. T. (1931). On the human remains excavated at Kish. Journal of the

Royal Anthropological hzstitute 61, 57-119. Cipriani, L. (1927). Su alcani crani Etruschi della Marsiliana. Studi Etruschi 1, 392-405. Crichton, J. M. (1966). A multiple diseriminant analysis of Egyptian and African Negro crania. Peabody

Museum Papers (Harvard University) 57, 47-67. Dawkins, W. B. (1900-1901). Skulls from cave burials at Zakro. Annual of the British School at Athens 7,

150-155. Deshayes, J . (1963). Rapport preliminaire sur les deux premieres campagnes de fouille A Tureng Tepe.

Syria 40, 85-99. Deshayes, J. (1973). Rapport preliminaire sur la neuvieme campagne de fouille a Tureng Tepe. ban 11,

141-152. Ehrich, R. W. (1939). Late cemetery crania. In (R. F. S. Starr) Nuzi L Cambridge: Harvard Univer-

sity Press. Finkel, D. J . (1974). Human remains at Tell Gezer. In (J. Seger, ed.) Gezer IV, The Field I Caves.

Jerusalem: Keter Publishing Co. Finney, D . J . (1971). Probit Analysis. Cambridge: Cambridge University Press. Furst, C. M. (1939). The skeletal material collected during the excavation of Dr. T. J. Arne in Shah

Tepe at Astrabadgorgan in Iran. The Sino-Swedish Expedition VIII (Archaeology) 9, 9-34. Gilles, E. & Elliot, O. (1963). Sex determinaion by discriminant function analysis. American Journal of

Physical Anthropology 21, 53-68. Howells, W. W. (1966). Craniometry and multivariate analysis. Peabody Museum Papers (Harvard

University) 57, 1-43. Hrdlicka, A. (1940). Skeletal remains. In (P. L. O. Guy & R. M. Engberg) Megiddo Tombs. Oriental

Institute Publications 33, 192-208. Kansu, S. A. (1937). Etude anthropologique de quelque squelettes d'Alacahuyuk. Belleten, Turk Tarih

Kurumu 1, 192-209. Kenyon, K. (1960). Archaeology in the Holy Land. London: Ernest Benn. Knussmann, V. R, (1968). Penrose-Abstand diskriminanzanalyse. Homo 18, 134-140.

DIMENSIONS OF M. E. SKELETAL POPULATIONS 229

Krogman, W. M. (1933). The cranial types. In (E. F. Schmidt) The Alishar Huyuk, Seasons of 1928 and 29, part 2, Oriental Institute Publications 20, 122-128.

Krogman, W. M. (1937). Cranial types from Alishar Huyuk and their relations to other racial types, ancient and modern, of Europe and Western Asia. In (H. H. van der Osten) The Alishar Huyuk, Seasons of 1930-32, part 3, Oriental Insitute Publications 30, 213-293.

Krogman, W. M. (1940). Racial types from Tepe Hissar, Iran, from the late fifth to the early second millenium B.C. Verhandelingen der Koninklijke Nederlandsche Akademie van Wetensehapben , Afdeeling Natuurkunde 39, 1-87.

Krogman, W. M. (1949). Ancient cranial types at Chatal Huyuk and Tell-al-Judaidah, Syria, from the late fifth millenium B.C. to the mid seventh century A.D. Belleten, Turk Tarih Kurumu 13, 407-477.

MacEwan, C. W. (1937). The Syrian expedition of the Oriental Institute of Chicago. American Journal of Archaeology 41, 8-16.

Morant , G. M. (1925). A study of Egyptian craniology from prehistory to Roman times. Biometrika I7, 1-52.

Risdon, D. L. (1939). A study of the cranial and other human remains from Palestine excavated at Tell Duweir (Lachish) by the Wellcome-Marston Archaeological Research Expedition. Biometrika 31, 99-166.

Schumacher, O. (1926). Uber alt-griesche schadel yon Myrina und Ephesus. Zeitschriftfur Morphologie und Anthropologie 25, 435-463.

Senyurek, M. S. (1952). A study of the human skeletons from Kultepe, excavated under the auspieces of the Turkish Historical Society. Belleten, Turk TarihKurumu 16, 323-343.

Sergi, G. (I 901 ). Notes upon the skulls of Erganos. American Journal of Archaeology 5, 315-328. Sokal, R. R. and F. J. Rohlf. (1969). Biometry. San Francisco: W. H. Freeman and Co. Swindler, D. R. (1956). A Study of the Cranial and Skeletal Material Excavated at Nippur, Museum Monographs.

Philadelphia: University of Pennsylvania Press. Talbot, P. A. & Mulhall, H. (1962). The Physical Anthropology of Southern Nigeria. Cambridge: Cam

bridge University Press. Tunakan, S. (1964). Bodrum-dirmil kasisis iskeletleri. Belleten, Turk Tarih Kurumu 28, 361-371. Woo, T. L. (1930). A study of 71 Ninth Dynasty Egyptian skulls from Sedment. Biometrika 22, 65-93. Wright, S. (1931). Evolution in Mendelian populations. Genetics 16, 97-159.